Measuring Chemistry Students' Understandings of Multiple External Representations through Cluster Analysis

学习化学和相关分子科学的一个重要方面是培养对分子、化合物、原子和离子的表征进行解释、生成和交流的能力。这个改进本科STEM教育（IUSE）研究项目将调查学生如何解释学生经常持有误解（或替代概念）的四个核心化学概念的表示：电子结构和周期性，分子形状和极性，键能和反应热，溶液化学和沉淀反应。基于对学生的深入访谈，研究小组将设计四个独立的调查来测试学生对这些主题的理解，并从全国范围内注册化学入门课程的本科生样本中收集数据。一旦这些工具被开发和验证，项目团队还将举办研讨会，教导其他化学教师和化学教育研究人员如何最好地利用这些工具来评估学生的理解，并提供基线数据，从而最终建立教学法和课程创新。本探索性研究项目将探讨以下问题：(1)本科生如何在临床访谈中解释核心化学概念（MERC3s）的多种外部表征？在宏观、微观和符号领域内和跨领域出现了什么认知失调？出现了什么替代概念？(2)在开发测量认知失调和另类概念的评估项目时，信度和效度是如何建立的？(3)通过对基于学生认知失调和MERC3s替代概念的评估项目反应的聚类分析，揭示了学生关于MERC3s的推理模式？为了确保研究结果的有效性和可靠性，项目团队将采用顺序、混合方法的研究设计，并对数据进行严格的心理测量测试，包括聚类分析。聚类分析也将使用现有的关于学生对其他核心化学概念的外部表征的理解的数据集进行。使用聚类分析来挖掘现有的和新的数据集将使研究团队不仅可以表征学生对单个核心概念表示的理解和使用，还可以表征学生在多个核心化学概念中概括的思维模式。此外，在项目的第二年和第四年，化学教育研究能力建设讲习班将重点向来自全国各地的100名研究生和博士后教授本研究中使用的先进统计技术，进一步推进学科教育研究的方法库。

An important aspect of learning about chemistry and related molecular sciences is developing abilities to interpret, generate, and communicate with representations of molecules, compounds, and the atoms and ions from which they are built. This Improving Undergraduate STEM Education (IUSE) research project will investigate how students interpret representations of four core chemistry concepts about which students often hold misconceptions (or alternative conceptions): electron structure and periodicity, molecular shape and polarity, bond energies and heats of reaction, and solution chemistry and precipitation reactions. Based on in-depth interviews with students, the research team will design four separate surveys to test students' understandings of these topics and to collect data from a national sample of undergraduate students enrolled in introductory chemistry courses. Once these tools have been developed and validated, the project team will also hold workshops to teach other chemistry instructors and chemistry education researchers how to best utilize the tools to assess students' understandings and provide baseline data upon which innovations in pedagogy and curriculum can ultimately be built. This exploratory research project will investigate the questions: (1) How do undergraduate students interpret multiple external representations of core chemistry concepts (MERC3s) within clinical interviews? What cognitive dissonance emerges, both within and across macroscopic, particulate, and symbolic domains? What alternative conceptions emerge? (2) How are reliability and validity best established when developing assessment items to measure cognitive dissonance and alternative conceptions? (3) What patterns in student reasoning about MERC3s are revealed through cluster analyses of responses to assessment items grounded in the students' cognitive dissonance and alternative conceptions about MERC3s? To ensure valid and reliable findings, the project team will use a sequential, mixed-methods research design and subject the data to rigorous psychometric testing, including cluster analysis. Cluster analyses will also be carried out with existing data sets regarding students' understanding of external representations of other core chemistry concepts. The use of cluster analyses to mine existing and new data sets will allow the research team to not only characterize students' understanding and use of representations of individual core concepts, but also characterize patterns in students' thinking that generalize across multiple core chemistry concepts. In addition, in years 2 and 4 of the project, chemistry education research capacity-building workshops will focus on teaching the advanced statistical techniques used in this research to a total of one hundred graduate students and postdoctoral scholars from across the country, further advancing the methodological repertoire within discipline-based education research.